| Message ID | 541C4C97.2010503@redhat.com |
|---|---|
| State | New |
| Headers | show |
Hi, On Fri, 19 Sep 2014, Aldy Hernandez wrote: > Michael, I really don't understand the need for this change in your original > patch. I don't know if this was a temporary testing change or what. I'm pretty sure it was temporary testing, when I still was finding my way through dwarf2out limitations/constraints. > I'm happy to report that with this and the last set of patches, both C > and C++ guality tests have <= regressions than mainline. Yay. Super. Ciao, Michael.
On 09/22/14 08:17, Michael Matz wrote: > Hi, > > On Fri, 19 Sep 2014, Aldy Hernandez wrote: > >> Michael, I really don't understand the need for this change in your original >> patch. I don't know if this was a temporary testing change or what. > > I'm pretty sure it was temporary testing, when I still was finding my way > through dwarf2out limitations/constraints. Ah perfect! > >> I'm happy to report that with this and the last set of patches, both C >> and C++ guality tests have <= regressions than mainline. Yay. > > Super. Again, thank you for your original patchset, which has immensely helped me navigate my way around the black box which was/is dwarf2out. Aldy
diff --git a/gcc/dwarf2out.c b/gcc/dwarf2out.c index f6c7f4a..68b4650 100644 --- a/gcc/dwarf2out.c +++ b/gcc/dwarf2out.c @@ -3683,7 +3683,7 @@ decl_ultimate_origin (const_tree decl) /* output_inline_function sets DECL_ABSTRACT_ORIGIN for all the nodes in the function to point to themselves; ignore that if we're trying to output the abstract instance of this function. */ - if (/*DECL_ABSTRACT (decl) &&*/ DECL_ABSTRACT_ORIGIN (decl) == decl) + if (DECL_ABSTRACT (decl) && DECL_ABSTRACT_ORIGIN (decl) == decl) return NULL_TREE; /* Since the DECL_ABSTRACT_ORIGIN for a DECL is supposed to be the
Michael, I really don't understand the need for this change in your original patch. I don't know if this was a temporary testing change or what. For instance, you had the following commented out in your original patch: if (/*DECL_ABSTRACT (decl) &&*/ DECL_ABSTRACT_ORIGIN (decl) == decl) return NULL_TREE; but then you also had: tree origin = decl_ultimate_origin (decl); <snip> <snip> - if (origin != NULL) + if (origin != NULL && origin != decl) It seems to me that if origin is not null, then "origin != decl" will always be true, since with your change, decl_ultimate_origin() will always return NULL when DECL_ABSTRACT_ORIGIN (decl) == decl. I suppose your /*DECL_ABSTRACT (decl) &&*/ change could be used in other uses of decl_ultimate_origin() that don't have this "&& origin != decl" business. Anyway... I don't understand the need for it (actually, to be honest, I don't understand what you were trying to do). I'm going to remove it. If you feel strongly about it, please chime in, and perhaps provide an explanation so I can include a comment. I'm happy to report that with this and the last set of patches, both C and C++ guality tests have <= regressions than mainline. Yay. Committing to mainline. Thanks. Aldy commit 7254ec0893f9b56a0ba78de9eab8895c9582c24b Author: Aldy Hernandez <aldyh@redhat.com> Date: Thu Sep 18 16:54:39 2014 -0600 * dwarf2out.c (decl_ultimate_origin): Allow checking of DECL_ABSTRACT.